The GAOLAOZHUANGREN1 Gene Encodes a Putative ...

5 downloads 0 Views 957KB Size Report
3 Cold Spring Harbor Laboratory, 1 Bungtown Road, Cold Spring Harbor, NY, U.S.A.. 4 Plant Biotechnology Center, Department of Horticulture and Crop ...
Plant Cell Physiol. 45(10): 1453–1460 (2004) JSPP © 2004

The GAOLAOZHUANGREN1 Gene Encodes a Putative Glycosyltransferase that is Critical for Normal Development and Carbohydrate Metabolism Minghai Shao 1, Huiqiong Zheng 1, Yi Hu 2, Dehua Liu 3, Jyan-Chyun Jang 4, Hong Ma 2, 3 and Hai Huang 1, 5 1

National Laboratory of Molecular Genetics, Shanghai Institute of Plant Physiology and Ecology, Shanghai Institute for Biological Sciences, Chinese Academy of Sciences, 300 Fenglin Road, Shanghai 200032, China 2 Department of Biology and the Huck Institute for Life Sciences, Pennsylvania State University, University Park, PA 16802, U.S.A. 3 Cold Spring Harbor Laboratory, 1 Bungtown Road, Cold Spring Harbor, NY, U.S.A. 4 Plant Biotechnology Center, Department of Horticulture and Crop Science, and Department of Plant Cellular and Molecular Biology, Ohio State University, Columbus, OH 43210, U.S.A. ;

Glycosyltransferases are enzymes that catalyze the attachment of a sugar molecule to specific acceptor molecules. These enzymes have been shown to play important roles in a number of biological processes. Whereas a large number of putative glycosyltransferase genes have been identified by genomic sequencing, the functions of most of these genes are unknown. Here we report the characterization of an Arabidopsis mutant, designated gaolaozhuangren1 (glz1), which is allelic to parvus characterized recently. The glz1 mutant exhibited a reduced plant stature, reduced size of organs in the shoot and dark-green leaves, indicating an important role of GLZ1 gene in normal development. The earliest GLZ1 expression appears at the shoot apical region of 4-d-old seedlings, which coincides with the onset of the glz1 morphological phenotypes. GLZ1 is expressed in a tissue-specific and developmentally regulated manner, predominantly in the stem and silique, and moderately in the flower. GLZ1 expression is strong in the midrib of rosette and cauline leaves; however, its expression was not detectable in the midrib of the cotyledon. Further analyses revealed that carbohydrate composition and distribution were aberrant in the glz1 mutant. These, together with the GLZ1 expression pattern, suggest a requirement for the GLZ1 function in normal sink–source transition during plant development. Keywords: Arabidopsis — Assimilate distribution — Development — Glycosyltransferases. Abbreviations: GLZ1, GAOLAOZHUANGREN1; Glc, glucose; Fru, fructose; Suc, sucrose.

Introduction Glycosyltransferases are involved in biochemical reactions that transfer sugar molecules to various acceptors in living cells. More than 7,200 glycosyltransferase genes have been 5

identified thus far in prokaryotes and eukaryotes (Coutinho et al. 2003). On the basis of amino acid sequence, glycosyltransferases have been classified into 65 families (Campbell et al. 1997, Coutinho et al. 2003). Plant glycosyltransferases play divergent roles in different physiological processes including glycolipid (Jarvis et al. 2000, Jorasch et al. 2000) and polysaccharide synthesis (Myers et al. 2000), as well as those relating to cell wall synthesis (Reiter et al. 1997, Falk et al. 2000, Perrin et al. 2001, Bouton et al. 2002, Iwai et al. 2002). In addition, plant glycosyltransferases are capable of adding sugars to small molecules, such as plant hormones (Szerszen et al. 1994, Martin et al. 1999a, Martin et al. 1999b, Jackson et al. 2001), salicylic acid (Lee and Raskin 1999) and flavonoids (Hirotani et al. 2000). In Arabidopsis, genomic sequencing and annotation have revealed more than 350 putative glycosyltransferase genes (Arabidopsis Genome Initiative 2000). However, the biochemical activities have only been determined for a handful of gene products. The difficulty in assessing the biochemical function of many glycosyltransferases may be in part due to the highly specific activities of these enzymes (Keegstra and Raikhel 2001). Genetic analysis of the quasimodo1 (qua1) mutations in Arabidopsis revealed an important function for the QUA1 gene, which encodes a putative membrane-bound glycosyltransferase (Bouton et al. 2002). Loss-of-function of QUA1 resulted in a 25% reduction of galacturonic acid levels in mutant cell walls, compared with those in the wild-type plants. This defect severely affected the development of qua1 mutant plants, resulting in a dwarf stature and numerous cell protrusions from leaves and hypocotyls (Bouton et al. 2002). These results indicate that QUA1 is required for normal pectin synthesis and cell adhesion. The QUA1 protein is a member of the glycosyltransferase family 8 (Bouton et al. 2002). This family comprises proteins of diverse organisms from bacteria to yeast, plant and human; their functions are largely unknown. Sink to source transition is a defining feature during leaf development. For instance, young leaves are heterotrophic (sink); they depend in part on carbohydrate imported from other regions of the plant. In contrast, mature leaves are

Corresponding author: E-mail, [email protected]; Fax, +86-21-64042385. 1453

1454

Role of GLZ1 in plant development

Fig. 1 Phenotypes of the glz1 mutant. The glz1 mutant shows a dramatic reduction in plant stature at mature stage (A). Five-dayold glz1 mutant and the wild-type seedlings do not show obvious size difference (B). Although overall architecture and organ shape are nearly normal, organ size in glz1 is reduced, including leaves (C), floral organs (D) and fruits (E). A top view of a flower shows the release of pollen from anthers in the wild type (F) but not in the glz1 mutant (G) grown at 28°C. Seedlings grown in soil for 7 d showing the size difference between wild type (H) and the glz1 mutant (I). Seedling morphology was nearly identical prior to the emergence of true leaves. In contrast to the wild type, glz1 cotyledon ceased to expand after the emergence of true leaves. WT, wild type Ler. Bars = 0.5 cm in (A), (C) and (D); bar = 1 cm in (B); and bars = 1 mm in (E), (F), (G), (H) and (I).

autotrophic (source); they produce excess amounts of assimilates, which are transported to sinks (Turgeon 1989). While it is known that glycosyltransferases are involved in complex carbohydrate metabolism, their effects on sink–source transition have not been revealed. Here we report the characterization of the Arabidopsis gaolaozhuangren1 (glz1) mutant and identification of the GLZ1 gene via an enhancer trap approach (Sundaresan et al. 1995). GLZ1 encodes a protein identical to PARVUS (Lao et al. 2003), which belongs to glycosyltransferase family 8. Because sugar was not translocated via a default pathway in the glz1 mutant, i.e. from source to sink, we propose that GLZ1 may play a role in sink–source transition by direct and indirect influence on carbohydrate metabolism, cell wall composition and transport function.

Results Isolation and characterization of the glz1 mutant To identify genes important for plant development, we screened Ds insertional lines for mutants with morphological defects. One Ds line was identified to segregate for mutant plants that showed a reduced plant stature and was named glz1.

The name gaolaozhuangren1 (glz1) came from a group of small people in a modern Chinese novel. Segregation analyses indicated that the glz1 phenotype was recessive. F1 progeny of a cross between glz1 and Ler all showed wild-type phenotypes, and the F2 generation segregated normal to mutant plants in a 3 : 1 ratio (331 : 107), indicating a single nuclear loss-of-function mutation. Compared with the wild type, the glz1 mutant was much smaller, with dark-green leaves (Fig. 1A). Although the glz1 mutant had nearly normal overall architecture and organ shapes, the size of all organs was reduced, including leaves (Fig. 1C), floral organs (Fig. 1D) and fruits (Fig. 1E). The size reduction of all organs was due to the reduction of both cell size and cell number (histological data not shown). In addition, the fertility of the glz1 mutant was poor and temperature sensitive. At 23°C, glz1 plants produced small siliques containing a few seeds each (Fig. 1E, middle), whereas at 28°C, glz1 plants were completely sterile (Fig. 1E, right). Reduced fertility was caused in part by a severe reduction of pollen grains in the glz1 mutant (Fig. 1F, G) at higher temperature. Interestingly, the morphology of young seedlings was indistinguishable between glz1 and wild type, including the size of cotyledons and the

Role of GLZ1 in plant development

1455

Fig. 2 Revertant sequencing and genetic complementation of the glz1 mutant. (A) Schematic representation of the Ds insertion, endogenous and transgenic GLZ1 genes and related primers (arrows) used for verifying complementation of the glz1 mutant. A triangle indicates the Ds element, and filled and empty boxes represent the GLZ1 coding region and CaMV 35S promoter, respectively. (B) Revertants with the complete wild-type phenotype were identified from progeny of glz1 mutant plants carrying an Ac element, and the Ds insertion sites of three revertants were sequenced. Rev#1 has the wild-type sequence. The underlined bases are from the duplication due to Ds insertion; and the boldfaced ones are extra bases added during the repair following Ds excision. (C) PCR detection of the transgenic GLZ1 gene in T1 plants using 35S and 2247-A primers (left panel). Lane P, plasmid containing the CaMV35S::GLZ1 construct; lane L, wild type Ler; lanes 1– 5, individual T1 transgenic plants. Genotypes of the transgenic T1 plants were determined by PCR (right panel). PCR experiments were performed with combined primers 2247-A, 2247-B, Ds3-3 and 2247-5. Lane M, homozygous glz1 mutant; lane L, wild type Ler; lanes 1–5, individual T1 transgenic plants. The PCR products, 5A, DA and BA (arrows) were amplified by primer combinations of 2247-5/ 2247-A, Ds3-3/2247-A and 2247-B/2247-A, respectively. (D–F) Complementation of the glz1 mutant. (D) A wildtype plant (positive control). (E) A T1 transgenic plant containing CaMV35S::GLZ1 in the glz1/glz1 background. (F) A homozygous glz1 mutant plant (negative control). Bars = 5 mm in (D), (E) and (F).

length of roots (Fig. 1B). The earliest difference in leaf size and leaf color was detected when the first pair of rosette leaves was expanding (Fig. 1I, H), at a time when photoassimilates are transported from the cotyledons to new leaves. These results suggest that GLZ1 function is not critical before the onset of true leaf development. GLZ1 encodes a putative glycosyltransferase The glz1 mutant contains a single Ds element (Southern results not shown) with a kanamycin resistance marker, which co-segregated with the glz1 mutant phenotypes (in total, 114 F2 lines with glz1 phenotypes were analyzed), indicating the link between the putative GLZ1 gene and the Ds insertion. The flanking sequence of the Ds insertion was obtained by using TAIL-PCR (Liu et al. 1995). A coding region (At1g19300) was identified by BLAST-searching of the Arabidopsis genome sequence using the Ds flanking sequence. Sequence analysis revealed that the Ds element was inserted into the transcribed region 36 bp upstream of the start codon of a gene (Fig. 2A), with a characteristic 8-bp duplication of genomic sequence flanking the Ds element. This gene encodes a protein in family 8 of the glycosyltransferase superfamily, identical to recently reported PARVUS (Lao et al. 2003).

To determine whether the glz1 mutation was caused by a Ds insertion, we screened for revertants and performed functional complementation by transforming glz1 mutant plants with the putative GLZ1 driven by CaMV 35S promoter (Fig. 2A). Revertants with complete wild-type phenotypes were identified from progeny of mutant plants carrying an Ac element. Sequence analysis of several revertants at the putative GLZ1 locus revealed footprints in two separate revertants (Fig. 2B), as well as wild-type sequences in others (Fig. 2B and data not shown). In the complementation experiment, five independent transgenic lines were obtained and all of them showed wild-type phenotypes. T1 and T2 plants that were in the glz1/ glz1 background and carried the CaMV35S::GLZ1 transgene all showed wild-type phenotypes (Fig. 2C–F). Fig. 2E shows a plant from line 1, with a copy of the CaMV35S::GLZ1 transgene in the glz1/glz1 background. Taken together, these results confirm that the Ds insertion into the GLZ1 gene is responsible for the glz1 phenotype. GLZ1 expression pattern To better understand GLZ1 function, we examined GLZ1 expression in different wild-type organs by RNA gel blot analyses. GLZ1 cDNA probe detected a 1.3-kb band predominantly in stem and young silique, moderately in flower and

1456

Role of GLZ1 in plant development

Fig. 3 Tissue-specific expression of GLZ1. RNAs were prepared from 4-week-old greenhouse soil-grown plants. The upper panel shows the expression using GLZ1 cDNA as a probe. ACTIN probe was used in the lower panel as control to adjust variation caused by RNA loading.

weakly in mature leaf. GLZ1 expression in roots was not detectable (Fig. 3). GLZ1 was not expressed in 4-week-old glz1 plants, indicating a block of transcription by Ds insertion (Fig. 3). The Ds insertion carried a GUS reporter gene when inserted into the 5′ untranslated region of the GLZ1 gene, providing a convenient system to monitor the spatial and temporal expression of GLZ1. To learn whether GLZ1 expression is developmentally regulated, we analyzed the patterns of GLZ1::GUS expression in plants heterozygous for the glz1 insertion. GUS staining was not detectable in most organs in young seedlings (date not shown). Undetectable GUS expression was consistent with the fact that the glz1 mutant at the equivalent developmental stage does not show any phenotype (Fig. 1B). The earliest appearance of clear GUS staining was at the shoot apical region (Fig. 4A, arrowhead) 4 d after germination. GUS staining became more distinct after the onset of true leaf devel-

opment. The GUS staining was intense in the class I vein (midrib) and trichomes, but weak in the class II vein (arising as single branches from the midrib) in nearly fully expanded leaves (Fig. 4B, C). Whereas GUS staining began to fade in other cell types in older leaves, midrib expression remained constant (Fig. 4D). The strongest GLZ1 expression was seen in inflorescence stems (Fig. 4E). Cross-sections showed that although GUS activity was detected in most cells, the highest expression appeared in the xylem, phloem and endodermis (Fig. 4F, G). In flower buds, GUS staining was visible in sepals (Fig. 4H); however, in mature flowers GUS expression only appeared in the stamen, filament and carpel (Fig. 4I). GUS activity was found in the septum in developing siliques (Fig. 4J). GUS staining was also detected in a region close to the shoot–root junction in 20-day-old plants (data not shown). This highly localized expression was not detectable in the root sample using RNA gel blot analysis. Altered carbohydrate composition in glz1 mutant Glycosyltransferases play pivotal roles in carbon metabolism in plants. To determine whether carbon metabolism in glz1 was affected, we conducted a simple comparison of carbohydrate composition between glz1 and wild-type leaves. In wildtype plants, starch content was markedly higher in the first pair of rosette leaves than in the second pair. However, there was no significant difference between the two pairs of rosette leaves in glz1. Furthermore, starch content of the second pair of leaves was slightly higher in glz1 than in the wild type (Fig. 5A). Soluble sugar content were also different between wild-type and mutant plants. Whereas glucose (Glc) and fructose (Fru) contents in the glz1 mutant were much higher in both the first and second pair of leaves than those in wild-type leaves, the

Fig. 4 Histochemical analysis of GLZ1 expression in the GLZ1/glz1 plants where no developmental defect was observed. In a 4-day-old seedling (A), GUS staining only appeared at the shoot apex (arrowhead). Young leaves usually showed more extensive GUS activity in the class I and class II veins (B) and trichomes (C). Mature leaves showed more restricted GUS activity in the midrib (D). Strong GLZ1 expression was detected in stems (E, F) where GUS staining was more localized in phloem (ph), xylem (xy) and endodermis (en) (G). In flower buds, GUS staining was visible in sepals (H). However, in mature flowers, GUS activity only appeared in the anther, filament and carpel (I). In siliques, GUS activity was found in septum (arrow). Bars = 1 mm in (A), (B), (D), (E), (H) and (J); 0.25 mm in (C); 40 µm in (F) and (G); and 0.5 mm in (I).

Role of GLZ1 in plant development

1457

Fig. 6 The pattern of sugar translocation in the glz1 mutant is different from that in the wild-type plant. [14C]Suc was fed to one of the cotyledons of 17-day-old plants. Translocation of [14C]Suc is indicated by the presence of gray areas on the X-ray film where dissected tissues were overlaid. The assay was repeated independently in three different plants and the results were consistent with what is shown here. (A) The dissected tissues and radioactivity image of wild-type and glz1 seedlings. Numbers below the figure indicate the dissected leaves and stem, which is indicated in (B). (B) Schematic representation of 17-day-old seedling: 1,2 represent the cotyledons; 3,4 and 5,6 represent the first and second pair of leaves, respectively; 7 represents the stem and root.

Fig. 5 Carbohydrate composition is abnormal in glz1 mutant leaves. Seventeen-day-old greenhouse grown plants were used in the analyses. (A) Starch content was significantly lower in M12. L12 and L34 represent the first and second pairs of wild-type rosette leaves, respectively. M12 and M34 represent the first and second pairs of rosette leaves of the glz1 mutant, respectively. (B) Soluble sugar contents are higher in glz1 mutant than in the wild type. Mean ± SE is given for three independent samples.

sucrose (Suc) content was higher in the second pair of leaves but lower in the first pair of leaves in the mutant (Fig. 5B). These results suggest that glz1 has altered carbohydrate metabolism or transport. Abnormal sucrose translocation in glz1 mutant Sugar translocation is a means of providing energy and structural elements not only for plant growth and development but also for the defense response to biotic and abiotic stresses. To determine whether sugar translocation is defective in glz1, we fed 17-day-old plants with [14C]Suc and monitored its distribution. When [14C]Suc was fed via a single cotyledon, it could only be found in sink tissues including stem, root, developing inflorescence and the youngest rosette leaf in the wildtype plants (Fig. 6A, upper panel). In glz1, however, [14C]Suc

was found not only in the cotyledon opposite the treated one but also every organ throughout the plant (Fig. 6A, lower panel). We also fed the first rosette leaf of 25-day-old plants with [14C]Suc. The distribution pattern of the [14C]Suc was very similar to the earlier results for both wild type and mutant (data not shown). These results indicate that either sucrose translocation is abnormal or sink–source transition is defective in the glz1 mutant.

Discussion We have identified a GLZ1/PARVUS gene via the characterization of a dwarf mutant glz1. GLZ1/PARVUS encodes a putative glycosyltransferase containing 351 amino acids. The GLZ1/PARVUS protein belongs to family 8 of glycosyltransferases; members of this family share a conserved Prodom domain that includes a DXD motif, which is thought to be involved in metal ion binding (Unligil and Rini 2000). Several members of this family have been identified as being involved in the synthesis of lipopolysaccharides [lipopolysaccharide galactosyltransferase (EC 2.4.1.44), lipopolysaccharide glycosyltransferase (EC 2.4.1.58)], glycogen [glycogenin glycosyltransferase (EC 2.4.1.186)] and inositol 1α-galactosyltransferase (EC 2.4.1.123, http://afmb.cnrs-mrs.fr/ CAZY/GT_8.html). In Arabidopsis, the GLZ1/PARVUS protein belongs to a subfamily with 41 members. The C-terminus of these proteins is highly conserved in a region encompassing

1458

Role of GLZ1 in plant development

the Prodom domains, whereas the N-terminus is much less conserved (Tavares et al. 2000). GLZ1 expression is under strict temporal and spatial control. In early seedling development, when cotyledons are expanding, nutrients needed for growth are provided mainly by the storage in the cotyledons, and long-distance sugar transport has not yet begun. At this stage, the mutant seedlings did not show an obvious phenotype and GLZ1 expression was undetectable. GLZ1 expression becomes detectable at the onset of emergence of the first pair of rosette leaves, which marks the beginning of assimilate transport from source (cotyledon) to sink (expanding rosette leaves). The timing of these events coincides with the appearance of the glz1 mutant phenotype. The glz1 phenotype becomes more severe after the first pair of rosette leaves have fully expanded and other rosette leaves begin to emerge. From this stage on, GLZ1 is strongly expressed in midribs of rosette leaves and cauline leaves, as well as inflorescence stems, especially in the vascular tissues. The GLZ1 expression pattern implicates its role in assimilate transport or sink–source transition. In glz1, carbohydrate composition was abnormal. In the wild type, starch content was higher in the first (source) than in the second (sink) pair of rosette leaves, presumably due to higher photosynthetic competence. However, starch content was approximately the same in both pairs of rosette leaves (sink) in the glz1 mutant and the concentration was comparable to that in the second pair of wild-type leaves. By contrast, the soluble sugar content was much higher in both pairs of glz1 leaves, except that sucrose content was lower in the first pair of glz1 leaves. We cannot rule out the possibility that sucrose concentration was underestimated in the first pair of glz1 leaves. Together these results suggest that starch accumulation is somehow less active in the first pair of glz1 leaves. Intriguingly, this cannot be interpreted solely by the accumulation of glucose and fructose in glz1 because the difference in starch content (Fig. 5A, approximately 150 nmol mg–1) was much greater than that in soluble sugar content (Fig. 5B, approximately 25 nmol mg–1). As a result, total starch + soluble sugar content in the first pair of leaves was higher in the wild type than that in glz1 mutant. This implies that the first pair of glz1 leaves are not as photosynthetically competent as the wild type and thus suggests a delay/block of transition from sink to source. Starch content showed less difference in the second pair of leaves between glz1 and the wild type presumably because both remained as sink tissue. This notion is consistent with the results of the sugar transport assay (Fig. 6A) in which both pairs of glz1 leaves and the second pair of wild-type leaves were indicated as sugar-importing sink tissues. It is unclear whether accumulation of soluble sugar in glz1 leaves is a result of altered carbohydrate metabolism and/or transport from other organs. Possibilities include abnormal invertase and sugar transporter activities associated with a defective cell wall (Sherson et al. 2003) or decreased usage of glucose/fructose substrate due to defective glycosyltransferase.

The [14C]Suc feeding experiment demonstrated that sucrose translocation was abnormal in glz1. In the wild-type plants, the fed [14C]Suc was clearly targeted to the sink: young leaves or the apical region where new organs are initiated. Normal sugar transport from source to sink is important not only for the growth of new organs but also for the removal of excess assimilates to relieve feedback inhibition of photosynthesis by sugars in source leaves. In glz1, however, [14C]Suc fed from either cotyledon or rosette leaf moved throughout the plant. This, together with low starch content in the first pair of rosette leaves, indicates a delay or block of transition from sink to source in the glz1 mutant. Growing leaves obtain carbon from two sources: import and photosynthesis. The carbon for protein synthesis in growing leaves is derived preferentially, but not exclusively, from photosynthesis, whereas imported sucrose is preferentially used in the synthesis of structural carbohydrates (Joy 1967, Dickson and Larson 1975, Ho and Shaw 1977). One possibility is that the GLZ1 gene was related to the synthesis of a certain carbohydrate structure, which may play an important role in leaf growth and expansion. Loss-offunction of the GLZ1 gene is likely to cause structural defects such as abnormal cell wall that directly or indirectly affect cell expansion and hence hinder sink–source transition. Bouton et al. (2002) proposed that all Arabidopsis members of the glycosyltransferase family 8 might be important for pectin biosynthesis. Pectins are a family of complex polysaccharides present in all plant primary cell walls and may have multiple functions in plant growth and development (Ridley et al. 2001). This is supported by recent characterization of parvus, a mutant allelic to glz1, in which mutation caused an altered cell wall carbohydrate composition and possibly abnormal pectin biosynthesis (Lao et al. 2003). In summary, we have shown that loss-offunction of GLZ1 causes abnormal carbohydrate composition and aberrant sugar translocation. While it is clear that glycosyltransferase is pivotal for plant growth and development, the precise biochemical and cellular functions of GLZ1/PARVUS remain to be elucidated.

Materials and Methods Plant materials and growth conditions Ds insertional lines in the Landsberg erecta (Ler) genetic background were generated as described previously (Sundaresan et al. 1995). Mutant and wild-type Ler plants were usually grown at about 22–24°C in a greenhouse with a photoperiod of 16 h light and 8 h dark. Plant growth on agar plates was according to our previous conditions under continuous illumination at 22°C (Li et al. 1999). Mutant plants were identified in the F3 generation following crosses between Ac and Ds plants. Heterozygous siblings of the glz1 mutant plants were allowed to mature and their seeds planted to determine segregation frequency. Normal and mutant plants segregated in an approximate 3 : 1 ratio, supporting the hypothesis that the mutant phenotype was cause by a single-gene recessive mutation.

Role of GLZ1 in plant development Morphological observations and detection of GUS activity Fresh tissues of wild-type and mutant plants were examined using a SZH10 dissecting microscope (Olympus, Japan), and photographed using a Nikon E995 digital camera (Nikon, Japan). Histochemical detection of GUS activity was carried out with 5-bromo-4-chloro-3indolyl β-D-glucuronic acid (X-Gluc) as a substrate. Plant tissues were placed in X-Gluc solution [750 µg ml–1 X-Gluc, 100 mM NaPO4 (pH 7), 3 mM K3F3(CN)6, 10 mM EDTA and 0.1% Nonidet-P40] under vacuum for 10 min at room temperature, then incubated overnight at 37°C for 1–3 d. Tissues were cleared with 95% ethanol, and viewed and photographed in 50% glycerol, using the SZH10 dissecting microscope. Molecular cloning and revertant analysis Genomic DNA was isolated as described previously (Huang et al. 1994). To isolate the GLZ1 gene, we obtained a 0.6 kb genomic fragment from the 3′ end of the Ds element, using the TAIL-PCR procedure (Liu et al. 1995). The TAIL-PCR was performed with the Dsspecific primers (Ds3-1, Ds3-2 and Ds3-3; Grossniklaus et al. 1998) and arbitrary degenerate primers (Liu et al. 1995). The GLZ1-specific primers were designed based on the sequence information of the TAIL-PCR product: 2247–3, 5′-AGAGCTTAAGAATCCAACGCGA3′; and 2247-5, 5′-AGGATCCATGT CCCAACATCTT-3′ (Fig. 2A). To determine whether the glz1 mutation was caused by a Ds insertion, we screened for revertants. No obvious somatic revertant sectors in the F2 progeny from a cross between glz1 mutant plants and an Ac line (Ac-1, Sundaresan et al. 1995). We then analyzed F3 progeny from individual F2 lines with glz1 phenotypes, and identified F3 plants with wild-type-like plants. We PCR-amplified and sequenced the genomic region flanking the Ds insertion from these wild-type-like plants, and found that the Ds element had been excised with footprints left in two of those lines (Fig. 2B), indicating that Ds insertion resulted in glz1 phenotypes. Complementation of the glz1 mutation DNA BLAST analysis revealed that there is no intron in the GLZ1 gene, so that the genomic DNA and 2247-3 and 2247-5 primers were used to amplify the entire GLZ1 coding region. To complement the glz1 mutations, the amplified DNA fragment containing the entire putative GLZ1 coding region from the wild-type Columbia ecotype was cloned into the vector pGEM7Z(+) between BamHI and SacI sites to yield the plasmid H451. After sequence verification, a HindIII– BamHI fragment containing a 35S promoter from the pBI121 vector was inserted into the H451, upstream of the GLZ1 fragment, resulting in the plasmid H452. The 35S::GLZ1 fragment was then released from H452 by HindIII and NsiI digestion and inserted into the HindIII and PstI sites of a binary T-DNA vector pCAMBIA1301 to yield the construct H453. This construct was then introduced into plants heterozygous for glz1 mutation by Agrobacterium-mediated transformation. Five independent T1 transgenic lines were obtained following hygromycin selection; they all appeared normal. Since the genotypes of the transgenic plants could be glz1/glz1, +/glz1 or +/+, additional PCR experiments were performed to further confirm the transgenic complementation of the glz1 mutant. The primer design is shown in Fig. 2A. First, the existence of the transgene was verified by primers 35S (5′GCTCCTACAAATGCCATCA-3′) and 2247-A (5′-GACTCGGAGTCATCGAT-3′). Then, four primers, consisting of three pairs (2247-5 and 2247-A; Ds3-3 [5′-GTTACCGACCGTTTTCATCC-3′] and 2247A; and 2247-B [5′-GGGCTTGGAATCTAATC-3′] and 2247-A), were combined in PCRs to analyze each transgenic line. Among these primers, 2247-5 and 2247-A can yield a band that serves as internal control. The PCR results showed that lines 2 and 4 are in the +/+ genotype, lines 1 and 5 in the +/glz1 genotype and line 3 in the glz1/glz1 geno-

1459

type (Fig. 2C). The wild-type phenotype of line 3 indicates that the 35S::GLZ1 can indeed complement the glz1 mutant. We also analyzed T2 segregating progeny of line 1, 3 and 5 transformants. Our results demonstrated that plants with the glz1 phenotype in T2 all lost the 35S: :GLZ1 transgene in the glz1/glz1 genotype (data not shown). RNA gel blot hybridization analysis Total RNA isolation, gel separation and RNA blotting were according to our previous protocols (Huang et al. 1995). Blots were made using Amersham Nylon N+ membranes and hybridization was performed as recommended by the manufacturer using 32P-labeled probes (Pharmacia, Sunnyvale, CA, U.S.A.). Hybridization and washing were conducted at a stringent temperature (65°C). Carbohydrate analysis Seventeen-day-old seedlings of wild type and glz1 mutant were used to detect the carbohydrate content. The first and second pairs of rosette leaves were collected in liquid nitrogen separately and lyophilized when used. The lyophilized leaves (2.5 mg each) were heated in boiling water in Eppendorf tubes for 20 min to destroy the endogenous enzymes. The leaves were then ground and mixed with 200 µl double distilled water in a vortex. The resulting samples were incubated at room temperature for 25 min with occasional vortexes, followed by centrifugation at 10,000×g for 10 min. The supernatants were used for soluble sugar assays with an enzymic method (Hampp et al. 1994). The absorbance was recorded by a microplate reader system (Sunrise, Phenix, Candler, NC, U.S.A.). Starch in pellets was determined following enzymic digestion as described previously (Obenland and Brown 1994). Three independent samples were measured to get the mean ± SE. [14C]Feeding The [14C]Suc feeding experiment was performed according to the method described previously (Gottwald et al. 2000), with modifications. Briefly, one cotyledon (from 17-day-old seedlings) or rosette leaf (from 28-day-old seedlings) from each plant to be fed was crimped with forceps, and 0.3 µl of [14C]Suc (0.2 mCi ml–1, Sigma, St. Louis, MO, U.S.A.) was placed at the crimped point on the cotyledon or rosette leaf. The drop was left for 2 h before the seedlings were dissected. Each leaf from a fed seedling was wrapped with the Saran Wrap and exposed to a phosphor screen (Molecular Dynamics, Boston, MA, U.S.A.) for 24–48 h before the screen was scanned.

Acknowledgments We thank C. Flanagan for contributing to the generation and initial analysis of Ds insertional lines; H. Choi, A. Debowski, M. Kim, S. Moran for assistance in mutant screens and preliminary studies; T. Mulligan, A. Omes and J. Wang for plant care; and D. Braun for helpful discussions and comments on this manuscript. This research was supported by grants from the Chinese Administration of Science and Technology (863), the Chinese National Science Foundation and the Shanghai Scientific Committee to H.H., supported by the Cold Spring Harbor Laboratory and the Pennsylvania State University and the grants from the US National Science Foundation and the Chinese National Science Foundation (30128014) to H.M., and grants from OARDC, OPBC, USDA to J.-C.J.

References Arabidopsis Genome Initiative (2000) Analysis of genome sequence of the flowering plant Arabidopsis thaliana. Nature 408: 796–815.

1460

Role of GLZ1 in plant development

Bouton, S., Leboeuf, E., Mouille, G., Leydecker, M.T., Talbotec, J., Granier, F., Lahaye, M., Hofte, H. and Truong, H.N. (2002). QUASIMODO1 encodes a putative membrane-bound glycosyltransferase required for normal pectin synthesis and cell adhesion in Arabidopsis. Plant Cell 14: 2577–2590. Campbell, J.A., Davies, G.J., Bulone, V. and Henrissat, B. (1997). A classification of nucleotide-diphospho-sugar glycosyltransferases based on amino acid sequence similarities. Biochem. J. 326: 929–942. Coutinho, P.M., Deleury, E., Davies, G.J. and Henrissat, B. (2003). An evolving hierarchical family classification for glycosyltransferases. J. Mol. Biol. 328: 307–317. Dickson, R.E. and Larson, P.R. (1975) Incorporation of 14C-photosynthate into major chemical fractions of source and sink leaves of cottonwood. Plant Physiol. 56: 185–193. Falk, A., Bar-Pelet, M., DeRocher, A.E., Zeng, W., Rerrin, R.M., Wilkerson, C., Raikhel, N.V. and Keegstra, K. (2000) Biochemical characterization and molecular cloning of an α-1, 2-fucosyltransferase that catalyzes the last step of cell wall xyloglucan biosynthesis in pea. J. Biol. Chem. 275: 15082–15089. Gottwald, J.R., Krysan, P.J., Young, J.C., Evert, R.F. and Sussman, M.R. (2000). Genetic evidence for the in planta role of phloem-specific plasma membrane sucrose transporters. Proc. Natl Acad. Sci. USA 97: 13979–13984. Grossniklaus, U., Vielle-Calzada, J.P., Hoeppner, M.A. and Gagliano, W.B. (1998). Maternal control of embryogenesis by MEDEA, a polycomb group gene in Arabidopsis. Science 280: 446–450. Hampp, R., Egger, B., Effenberger, S. and Einig, W. (1994). Carbon allocation in developing spruce needles. Enzymes and intermediates of sucrose metabolism. Physiol. Plant. 90: 299–306. Hirotani, M., Kuroda R., Suzuki, H. and Yoshikawa, T. (2000) Cloning and expression of UDP-glucose:flavonoid 7-O-glucosyltransferase from hairy root cultures of Scutellaria baicalensis. Planta 210: 1006–1013. Ho, L.C. and Shaw, A.F. (1977) Carbon economy and translocation of 14C in leaflets of seventh leaf of tomato during leaf expansion. Ann. Bot. 41: 833– 848. Huang, H., Tudor, M., Weiss, C.A., Hu, Y. and Ma, H. (1995). The Arabidopsis MADS-box gene AGL3 is widely expressed and encodes a sequence-specific DNA-binding protein. Plant Mol. Biol. 28: 549–567. Huang, H., Weiss, C.A. and Ma, H. (1994). Regulated expression of the Arabidopsis Ga gene GPA1. Int. J. Plant Sci. 155: 3–14. Iwai, H., Masaoka, N., Ishii, T. and Satoh, S. (2002) A pectin glucuronyltransferase gene is essential for intercelluar attachment in the plant meristem. Proc. Natl Acad. Sci. USA 99: 16319–16324. Jackson, R.G., Lim, E.-K., Li, Y., Kowalczyk, M., Sandberg, G., Hoggett, J., Ashford, D.A. and Bowles, D.J. (2001). Identification and biochemical characterization of an Arabidopsis indole-3-acetic acid glucosyltransferase. J. Biol. Chem. 276: 4350–4356. Jarvis, P., Dormann, P., Peto, C.A., Lutes, J., Benning, C. and Chory, J. (2000) Galactolipid deficience and abnormal chloroplast development in the Arabidopsis MGD synthase 1 mutant. Proc. Natl Acad. Sci. USA 97: 8175–6179. Jorasch, P., Warnecke, D.C., Lindner, B., Zahlinger, U. and Heinz, E. (2000) Novel processive and nonprocessive glycosyltransferase from Staphyloccus aureus and Arabidopsis thaliana synthesize glycoglycerolipids, glucophospholipids, glycosphingolipids and glycosterols. Eur. J. Biochem. 267: 3770– 3783. Joy, K.W. (1967) Carbon and nitrogen sources for protein synthesis and growth in sugar-beet leaves. J. Exp. Bot. 18: 140–150.

Keegstra, K. and Raikhel, N. (2001). Plant glycosyltransferases. Curr. Opin. Plant Biol. 4: 219–224. Lao, N.T., Long, D., Kiang, S., Coupland, G., Shoue, D.A., Carpita, N.C. and Kavanagh, T.A. (2003) Mutation of a family 8 glycosyltransferase gene alters cell wall carbohydrate composition and causes a humidity-sensitive semisterile dwarf phenotype in Arabidopsis. Plant Mol. Biol. 53: 647–661. Lee, H. and Raskin, I. (1999) Purification, cloning, and expression of a pathogen inducible UDP-glucose:salycylic acid glucosyltransferase from tobacco. J. Biol. Chem. 274: 36637–36642. Li, F., Flanagan, C.A., Zhao, Y., Ma, H. and Huang, H. (1999). Assignment of 44 Ds insertions to the linkage map of Arabidopsis. Plant Mol. Biol. Rep. 17: 109–122. Liu, Y., Mitsukawa, N., Oosumi, T. and Whittier, R.F. (1995). Efficient isolation and mapping of Arabidopsis thaliana T-DNA insert junctions by thermal asymmetic interlaced PCR. Plant J. 8: 457–463. Martin, R.C., Mok, M.C. and Mok, D.W.S. (1999a). A gene encoding the cytokinin enzyme zeatin-O-xylosyltransferase of Phaseolus vulgaris. Plant Physiol. 120: 553–557. Martin, R.C., Mok, M.C. and Mok, D.W.S. (1999b). Isolation of a cytokinin gene, ZOG1, encoding zeatin-O-glucosyltransferase from Phaseolus lunatus. Proc. Natl Acad. Sci. USA 96: 284–289. Myers, A.M., Morell, M.K., James, M.G. and Ball, S.G. (2000) Recent progress toward understanding biosynthesis of the amylopectin crystal. Plant Physiol. 122: 987–997. Obenland, D.M. and Brown, C.S. (1994). The influence of altered gravity on carbohydrate metabolism in excised wheat leaves. J. Plant Physiol. 144: 696– 699. Perrin, R.M., Wilkerson, C. and Keegstra, K. (2001) Golgi enzymes that synthesize plant cell wall polysaccharides: finding and evaluating candidates in the genomic era. Plant Mol. Biol. 47: 115–130. Reiter, W.D., Chapple, C. and Somerville, C.R. (1997) Mutants of Arabidopsis thaliana with altered cell wall polysaccharide composition. Plant J. 12: 335– 345. Ridley, B.L., O’Neill, M.A. and Mohnen, D. (2001) Pectins: structure, biosynthesis, and oligogalacturonide-related signaling. Phytochemistry 57: 929–967. Sherson, S.M., Alford, H.L., Forbes, S.M., Wallace, G. and Smith, S.M. (2003) Roles of cell-wall invertases and monosaccharide transporters in the growth and development of Arabidopsis. J. Exp. Bot. 54: 525–531. Sundaresan, V., Springer, P., Volpe, T., Haward, S., Jones, J.D.G., Dean, C., Ma, H. and Martienssen, R. (1995). Patterns of gene action in plant development revealed by enhancer trap and gene trap transposable elements. Genes Dev. 9: 1797–1810. Szerszen, J.B., Szczyglowski, K. and Bandurski, R.S. (1994) iaglu, a gene from Zea mays involved in conjugation of growth hormone indole-3-acetic acid. Science 265: 1699–1701. Tavares, R., Aubourg, S., Lecharny, A. and Kreis, M. (2000). Organization and structural evolution of four multigene families in Arabidopsis thaliana: AtLCAD, AtLGT, AtMYST and AtHD-GL2. Plant Mol. Biol. 42: 703–717. Turgeon, R. (1989). The sink–source transition in leaves. Annu. Rev. Plant Physiol. Plant Mol. Biol. 40: 119–138. Unligil, U.M. and Rini, J.M. (2000). Glycosyltransferase structure and mechanism. Curr. Opin. Struct. Biol. 10: 510–517.

(Received March 19, 2004; Accepted July 22, 2004)